R E S E A R C H C O M M U N I C A T I O N
Anti-neuroinflammatory effect of daidzein in human
hypothalamic GnRH neurons in an in vitro
membrane-based model
Sabrina Morelli
1|
Antonella Piscioneri
1|
Giulia Guarnieri
2|
Annamaria Morelli
2|
Enrico Drioli
1,3|
Loredana De Bartolo
11Institute on Membrane Technology,
National Research Council of Italy, ITM-CNR, Rende, Italy
2Department of Experimental and Clinical
Medicine, University of Florence, Florence, Italy
3WCU Energy Engineering Department,
Hanyang University, Seoul, Republic of Korea
Correspondence
Sabrina Morelli and Loredana De Bartolo, Institute on Membrane Technology, National Research Council of Italy, ITM-CNR, Rende Italy.
Email: s.morelli@itm.cnr.it; sabrina. morelli@cnr.it (S. M.) and l.debartolo@ itm.cnr.it; loredana.debartolo@cnr.it (L. D. B.)
Funding information
Ministero dell'Università e della Ricerca, Italy (MIUR) - PRIN 2017, Grant/Award Number: 2017TK7Z8L_006
Abstract
Phytoestrogens can control high-fat diet-induced hypothalamic inflammation that is associated with severe consequences, including obesity, type 2 diabetes, cardiovascular and neurodegenerative diseases. However, the phytoestrogen anti-neuroinflammatory action is poorly understood. In this study, we explored the neuroprotection mediated by daidzein in hypothalamic neurons by using a membrane-based model of obesity-related neuroinflammation. To test the daidzein therapeutic potential a biohybrid membrane system, consisting of hfHypo GnRH-neurons in culture on PLGA membranes, was set up. It served as reliable in vitro tool capable to recapitulate the in vivo structure and func-tion of GnRH hypothalamic tissue. Our findings highlighted the neuro-protective role of daidzein, being able to counteract the palmitate induced neuroinflammation. Daidzein protected hfHypo GnRH cells by down-regulating cell death, proinflammatory processes, oxidative stress, and apopto-sis. It also restored the proper cell morphology and functionality through a mechanism which probably involves the activation of ERβ and GPR30 recep-tors along with the expression of GnRH peptide and KISS1R.
K E Y W O R D S
daidzein, human GnRH neurons, hypothalamic neurons, neuoroinflammation, PLGA membrane system
1
|
I N T R O D U C T I O N
Recent interdisciplinary studies in neuroscience have found a close relationship between overnutrition, espe-cially high-fat diets, and inflammation in the brain, and
particularly in the hypothalamus. The roinflammation causes a progressive damage of the neu-rons that control food intake and energy homeostasis, thus, leading to obesity, diabetes, and cardiovascular complications. Overnutrition-induced hypothalamic
Abbreviations: DCA, dynamic contact angle; DCF, dichlorofluorescein; DMSO, dimethyl sulfoxide; DZ, daidzein; ECM, extracellular matrix; ER, estrogen receptor; GnRH, gonadotropin releasing hormone; GPR-30, G-protein coupled receptor 30; hfHypo, human fetal hypothalamic cells; IL, Interleukin; KISS1R, kisspeptin receptor; LSCM, Laser Scanning Confocal Microscopy; MTT, 3-[4, 5-dimethylthiazol-2-yl]-2, 5-diphenyltetrazolium bromide; NeuN, neuronal nuclei; PA, palmitate; PLGA, Poly(D,L-lactide-co-glycolide); PSCDs, Polystyrene culture dishes; ROS, reactive oxygen species; SEM, scanning electron microscope; SFAs, saturated fatty acids.
DOI: 10.1002/biof.1701
inflammation is also linked to neurodegenerative dis-eases such as Alzheimer's disease, Parkinson's disease, and stroke.1Therefore, the investigation of potential anti-inflammatory agents that protect the hypothalamus from diet-induced inflammation represents an active area of research in the biomedical field.
Accumulating evidence suggests that intracellular dis-turbances and stresses including endoplasmic reticulum (ER) stress, oxidative stress, and autophagic defects are important mediators of hypothalamic inflammation. How-ever, it is still needed to highlight the multiple mechanisms implicated in the induction of obesity-related neu-roinflammation. Studies have demonstrated that elevated levels of saturated fatty acids (SFAs), such as the highly prevalent palmitate (PA), induce a state of neu-roinflammation, characterized by several metabolic and molecular changes and severe damages in hypothalamic neurons. The accumulation of PA and its metabolites in the hypothalamus has been shown to increase the mRNA expression of pro-inflammatory cytokines, chemokines, and reactive oxygen species (ROS), and can result in apo-ptotic cell death.2,3Nevertheless, the effect of PA-associated hypothalamic inflammation on the different population of hypothalamic neurons is not fully understood.
The finding that estrogenic compounds provide neuroprotection against neuroinflammation although they have many adverse effects (e.g., increased risk of uterine, breast and endometrial cancer, heart disease and liver complications) raise the need to explore substitute compounds that have estrogenic properties without toxic-ities, as an alternative to estrogen replacement therapy.
Isoflavones are natural polyphenolic compounds, which act as phytoestrogens that have a diphenolic struc-ture similar to estrogen with extensive pharmacological properties4; recently, their role as anti-obesity agents is emerging. Daidzein, 40,7-dihydroxyisoflavone, a major isoflavone from soybean, is one of the most studied and potent phytoestrogens.5,6 It has several activities such as antioxidant, anti-inflammatory, antitumor, and cardi-oprotective properties.7–9 Previous studies have also reported the neuroprotective actions of daidzein against hypoxia in primary cortical neurons, excitotoxicity, β-amyloid, stroke-like injury, and ER stress-mediated neuronal degeneration,10–12besides to axonal outgrowth in neuronal cells. Daidzein regulates pro-inflammatory gene expression by activating PPAR-α and -γ and inhibiting the JNK pathway in adipocyte and macro-phage co-cultures.13 Recent evidences show the neuro-protective effect of daidzein in hypothalamic N42 cells against palmitic acid induced oxidative stress and apopto-sis mediated by the transcriptional coactivator peroxi-some proliferator-activated receptorγ coactivator 1 alpha (PGC1-alpha), which plays a role in oxidative defense.14
Despite these data, the exact mechanism of daidzein's neuroprotective effect has yet to be fully elucidated, spe-cifically, the anti-neuroinflammatory action of daidzein in hypothalamic neurons is poorly understood and require further investigation. Indeed, considering the prominent role played by the fatty acids excess (palmitate administration) in the neuroinflammation induction2,3 and the daidzein mediated wide protection spectrum,7–14 it is reasonable to seek and highlight its capacity to stop and/or revers the inflammatory cascade.
To address this issue, the present work aims to eluci-date the potential protective role of daidzein against palmitate-induced impairment in hypothalamic neurons that were cultured in a biohybrid membrane system in order to create an in vitro model of obesity-related neuroinflammation.
Previous works provided a proof of concept that semi-permeable polymeric membranes are advanced biomate-rials with a feature complexity, mimicking the native extracellular matrix (ECM), and topographical, chemical, and bio-functional cues that enable neuronal growth and differentiation.15Membranes, differently from other bio-materials that have been used to support neuronal cells cultures (e.g., electrospun, hydrogels, self-assembling pep-tides, nanoparticles, polystyrene dishes) display highly selective properties, which allow to create a fully con-trolled microenvironment at molecular level. These sys-tems provide compartmentalization of cells governing the mass transfer of molecules to and from cells and offering appropriate structural and physico-chemical surface stim-uli at nano and micro-scale, which are important for suc-cessfully supporting and guiding neuronal tissue repair and regeneration.16–18Membrane-based approaches have been used for the realization of valuable in vitro brain tis-sue models as investigational platforms for studying molecular mechanisms in physiological and/or pathologi-cal states including neurodegeneration,19 drug testing,20 and potential therapeutic strategy.21,22
However, there is still a gap regarding the develop-ment of an in vitro human neural tissue model. In the current study, we realized a novel highly selective mem-brane system for the culture of human hypothalamic neurons. We employed a primary culture of human fetal hypothalamic gonadotropin-releasing hormone (GnRH) neurons (hfHypo), previously isolated, and characterized, as cellular model whose functions have shown to be impaired by inflammation.23 GnRH neurons are central regulators of the reproductive function and respond to inflammatory changes, thereby representing a crucial hypothalamic population implicated in the pathogenic mechanisms linking obesity to reproductive disorders.24 As previously shown, hfHypo cell function is deeply affected by metabolic syndrome/obesity-related neuroinflammatory
cytokines, such as TNFα,23 therefore, they represent a novel and valuable tool for investigating the direct effects of metabolic inflammatory factors in the human hypo-thalamus and to unravel the potential protective role of molecules, such as daidzein.
GnRH cell lines and other kinds of non-human immortalized cell lines have been used in some papers to study the effects of leptin, insulin and palmitate in the hypothalamus,25,26but the use of human primary GnRH neurons is limited. Moreover, there are no reports con-cerning the effects of daidzein on palmitate-induced neu-roinflammation in hypothalamic neurons.
Hence, for the first time, we explored the anti-neuroinflammatory effects and mechanisms of action of daidzein in hfHypo neurons treated with palmitate. In this regard, we hypothesized that dadzein, thanks to its steroid structure, might activate, and exert neuroprotective effects in several settings that involve activation of estrogen receptors, as well as in regulating specific markers of neu-ronal differentiation and/or GnRH phenotype. We investi-gated its efficacy in promoting cell survival, in attenuating and/or suppressing palmitate-induced ROS production, mitochondrial dysfunction and apoptosis, and in retaining GnRH phenotype in vitro, in order to explore the possible underlying molecular mechanism.
The overall application represents also a promising resource to use membrane technology as preliminary drug screening in vitro tool for identifying new therapies to treat neurological disorders as well as for preclinical screening alternatively to animal experimentation.
2
|
M A T E R I A L S A N D M E T H O D S
2.1
|
Membrane system: Preparation and
characterization
Membranes of Poly(D,L-lactide-co-glycolide) (PLGA) were prepared in flat configuration by phase inversion technique. PLGA (7,5% wt/vol) (Resomer® RG 505, Mn 54.000–69.000, Sigma-Aldrich, Italy) was dissolved in dichloromethane. The polymeric solution was cast uniformly on the surface of a glass plate by using a handle-casting knife (Elcometer, gap set at 250μm), dried at room temperature until com-plete solvent evaporation, and washed with distilled water. Then, the obtained PLGA membranes were characterized in order to evaluate their morphological, physico-chemical and mechanical properties that play an important role in guiding neuronal behavior.
Morphological properties of the PLGA membranes were characterized by scanning electron microscope (SEM). Sam-ples of membrane surface and cross section were mounted with double-faced conductive adhesive tape, and, after gold
coating, analyzed by SEM (ESEM FEG QUANTA 200, FEI Company) in order to evaluate surface and cross section structure, shape and sizes of the membrane pores.
The membrane thickness was measured with Carl Mahr 40E digital micrometer. Membrane mean pore diameter was determined using a Capillary Flow Por-ometer (CFP 1500 AEXL, Porous Materials Inc., PMI).
The physico-chemical properties of PLGA membrane ware characterized by water dynamic contact angle (DCA) measurements at room temperature by using a contact angle meter, CAM 200 (KSV Instruments, Ltd.).
Mechanical properties were evaluated by Zwick/Roell Z2.5 tensile testing machine. Membrane samples were cut into strips of 5× 1 cm, mounted between two pneumatic grips, and subjected to uniaxial tension until failure. Ulti-mate tensile strength (UTS), Young modulus (Emod) and Elongation at break parameter (ε) were determined by analyzing five replicates of membrane sample.
2.2
|
Cell culture and treatments
Hypothalamic GnRH neurons (hfHypo) were isolated from the brain of 12-week-old human fetuses as previ-ously described,23 and then cryopreserved. The use of human fetal biopsies for research purposes was approved by the National Ethics Committee and the local Ethics Committee for investigation in Humans of the University of Florence (Protocol Number: 678304) and was per-formed with the inper-formed consent signed by the donors.
In the present work, hfHypo at passages P12–13 were seeded at a density of 16× 103cell/cm2on PLGA mem-branes in Coon's modified Ham's F12 medium (Euroclone, Milan, Italy) supplemented with 10% fetal bovine serum, and cultured under standard conditions (37C, 5% CO2). Polystyrene culture dishes (PSCD) were
used as reference substrates. The previous work has shown the maintenance of the phenotype of this cell pop-ulation until the passage P23.23
To realize an in vitro model of obesity-related neu-roinflammation, neuronal cells after 48 h of culture in PLGA membrane system were exposed to sodium palmi-tate (PA, Sigma-Aldrich) at different concentrations (0–600 μM) for 24–48 h at 37C. Palmitate was dissolved in water (50C) to have a stock solution of 10 mM concentra-tion and diluted in culture medium. The possible toxic effect of daidzein was evaluated by treating cells with differ-ent concdiffer-entrations of molecule ranging from 0.1 to 10μM.
To investigate the neuroprotective effects of daidzein (DZ, Santa Cruz Biotechnology) against palmitate-induced toxicity, cells were treated with a mixture of palmitate (300μM) and daidzein (PA + DZ) at different concentrations (0.1, 1, and 10 μM) for 24 h at 37C.
Daidzein was dissolved in DMSO to have a 10 mM stock solution and added to the culture medium at the indi-cated concentrations. We used the nomenclature “Con-trol” to indicate untreated cells without palmitate and daidzein treatment. After 24 h under the different treat-ments, cells were sacrificed for morphological studies, and assessment of cell viability, neuronal behavior and functional responses, and the neuronal cell medium supernatant was collected and stored at−70C for further analyses. All experiments were conducted in triplicate with the average of five samples for treatment (Control, Palmitate and PA + DZ).
2.3
|
SEM analysis
The morphology of hfHypo-GnRH neurons was investi-gated by SEM analysis. All treated and untreated samples of neurons grown in PLGA membrane system were fixed with 2.5% glutaraldehyde for 30 min, followed by post-fixation in 1% osmium tetroxide and then dehydrated using a growing series of ethanol dilutions (10%–100%) and finally air-dried. Samples were examined at SEM and representative images displaying neuronal structural fea-tures were obtained.
2.4
|
Immunostaining for laser scanning
confocal microscopy
The behavior of hfHypo cells was analyzed by Laser Scanning Confocal Microscopy (LSCM, Fluoview FV300, Olympus) after the immunostaining of neuronal markers (β-Tub III and Neuronal nuclei, NeuN), GnRH pheno-type markers (GnRH peptide and kisspeptin receptor, KISS1R) and estrogenic receptors (ER β and G protein-coupled receptor 30, GPR30). Neurons were fixed in 4% (wt/vol) paraformaldehyde for 30 min, followed by permeabilization and blocking with a solution containing 0.3% (vol/vol) Triton X-100 and 10% (vol/vol) FBS in PBS for 1 h at 37C. Samples were then incubated overnight at 4C with the following primary antibodies: anti-βTub III (1:500, Covance), NeuN (1:200, Millipore), anti-GnRH (1:100, Abcam), anti-KISS1R (1:200, Santa Cruz Biotechnology), anti-ERβ (1:200, Santa Cruz Biotechnol-ogy) and GPR30 (1:200, Abcam). Secondary anti-bodies, Cy2TM-conjugated Affini Pure donkey anti-rabbit IgG, Cy3TM-conjugated Affini Pure donkey anti-mouseIgG and a Cy5TM-conjugated Affini Pure donkey anti-goat IgG (1:500, Jackson ImmunoResearch Europe Ltd.) were then added for 1 h at RT. Finally, cells were counterstained with 200 ng/ml DAPI (Molecular Probes), for nuclear localization.
2.5
|
Measurement of cell viability
Cell viability was measured by 3-(3,4-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide (MTT) assay (Sigma-Aldrich). After 24 h of treatments, the culture medium was replaced with serum-free medium con-taining 5 mg/ml of MTT and incubated for 2 h at 37C. Then, lysis buffer (10% SDS, 0.6% acetic acid in DMSO, pH 4.7) was added to dissolve formazan crystals, and the absorbance was measured with a spectrophotometer at a 570 nm wavelength. Cell viability is reported as percent-age relative to control (100%).
2.6
|
Cytokine determination
Cytokine production was assessed by analyzing the con-centration of specific interleukins: Interleukin-6 (IL-6) and Interleukin-10 (IL-10), which were quantified by per-forming enzyme-linked immunosorbent assay (ELISA) in the supernatants of neuronal cultures.
Specific detection of IL-6 and IL-10 concentrations were obtained by a quantitative sandwich enzyme immu-noassay technique.
A specific IL-6 ELISA kit (AbFrontiers) was employed for the quantitative detection of the IL-6 released by cells in the culture media, by following the manufacturer's protocol.
Monoclonal antibody specific to IL-10 (Thermo Scien-tific) was pre-coated onto a 96-well plate. Standards and cell culture supernatant were added to the wells with biotin-conjugated monoclonal antibody specific for IL-10. Therefore, all IL-10 was captured by the immobilized anti-bodies. Following a wash to remove any unbound sub-stances, horseradish peroxidase-streptavidin conjugate was added to bind to the biotin on IL-10. Following another wash, a substrate buffer containing Tetramethylbenzidine (Thermo Scientific) was added to the wells. The reaction was stopped with 8 N H2SO4. Absorbance was measured at
450 nm using a Multiskan Ex (Thermo Lab Systems) and was proportional to the amount of IL-10.
2.7
|
Cellular ROS detection assay
ROS generation in cells was measured using the probe 20,70-dichlorodihydrofluorescein diacetate (H2DCF-DA,
Sigma-Aldrich), as previously described.20This molecule in presence of ROS is deacetylated and oxidized becom-ing the green fluorescent dichlorofluorescein (DCF). Then an increase in DCF fluorescence intensity is propor-tional to the ROS level. Neuronal cells (16× 103cell/cm2) in the membrane system, after 24 h of treatment with
palmitate and the different concentration of daidzein, were loaded with 50μM H2DCF-DA at 37C for 30 min.
Then cells were washed twice with Hank's salt solution and the DCF fluorescence intensity was measured with LSCM, using an argon laser. Quantitative analysis was performed on confocal images of cells by using Fluoview 5.0 software (Olympus Corporation).
2.8
|
Mitochondrial membrane potential
measurement
The mitochondrial membrane potential of neurons was measured by using the fluorescent cyanine dye 5,50,6,60 -tetrachloro-1,10,3,30-tetraethylbenzimidazolcarbocyanine iodide (JC-1) (Sigma-Aldrich). JC-1 probe shows potential-dependent accumulation in mitochondria. An emitted red fluorescence (aggregates) indicates healthy mitochondria with normal potential, whereas the mito-chondria with reduced potential emit green fluorescence (monomers). Therefore, the ratio of red to green fluores-cence of JC-1 represents the condition of mitochondrial membrane potential. Cell cultures were incubated at 37C with JC-1 (5 μg/ml) for 20 min. The fluorescence intensity of both monomers and J-aggregates were detected by LSCM. JC-1 green fluorescence was visual-ized with Ar laser and red fluorescence from JC-1 aggre-gates was detected using He/Ne green laser. Then, red/green fluorescence intensity was measured, and a quantitative analysis was performed on confocal images of cells by using Fluoview 5.0 software.
2.9
|
Apoptosis assay
The apoptotic degree was evaluated by investigating the expression of two apoptotic markers, caspase-3 and phospho-JNK (p-JKN), which were detected by immuno-staining with rabbit anti-caspase-3 antibody (1:250, BD) and mouse anti-p-JNK (Thr183/Tyr 185) antibody (1:250, Santa Cruz Biotechnology), respectively. Quantitative analysis of apoptotic cells was performed by calculating the percentage of positive cells for each marker (caspase-3 positive nuclei and p-JNK positive nuclei) over the total (DAPI-stained nuclei).
2.10
|
Statistical analysis
Statistical significance of the experimental results was assessed using ANOVA test followed by Bonferroni t-test. The results were expressed as mean ± SEM from at least three independent experiments in which five samples for
each treatment were assessed. A p value <0.05 was con-sidered to be statistically significant.
3
|
R E S U L T S
3.1
|
PLGA membrane system sustains
the growth of hfHypo neurons
PLGA membrane underwent to a full characterization in order to disclose their intrinsic properties that strongly impact on hypothalamic cell behavior in reorganizing themselves according to a proper structure. For a faithful tissue-like assembly, the appropriate cell displacement must be reproduced as a consequence of the correct interactions between cells and with the surrounding environment.
Flat PLGA membranes were morphologically charac-terized through SEM analysis revealing a homogenous and smooth surface, with nanopores regularly distributed over it (Figure S1).
Porometer measurements pointed out a mean pore size distribution of 20 ± 2 nm (Table 1). Membrane phys-ical chemphys-ical properties directly influence cell-material interaction and therefore the overall tissue formation process. To better predict and evaluate membrane perfor-mance in terms of cell growth supporting ability, dynamic contact angle measurements were carried out. Surface characterization pointed out an advancing con-tact angle of 81± 3 and a receding one with a value of 44± 4, evidencing a moderate wettability which is a desirable condition for promoting an optimal interaction at the cellular-material interface. When cells are cultured in vitro they can sense material rigidity, exhibiting a stiffness-dependent behavior on the basis of the mechani-cal features that characterize their natural niche. Biome-chanical stimuli arising by the substrate are responsible of cell migration, polarization, and differentiation, rep-resenting therefore an important membrane characteristic
T A B L E 1 Poly(D,L-lactide-co-glycolide) (PLGA) membrane properties
PLGA membrane
Thickness [μm] 9.3 ± 1
Mean pore diameter [nm] 20 ± 2
WCA [] θadv81 ± 3
θrec44 ± 4
E [N/mm2] 315 ± 17
ε [%] 11 ± 0.8
UTS [N/mm2] 7 ± 0.6
Note: Advancing θadvand recedingθreccontact angle; Young's modulus E;
that needs to be investigated. PLGA membrane exhibits a moderate stiffness as demonstrated by the Young's modu-lus value of 315 ± 17 N/mm2(see Table 1) in comparison with other materials utilized for neuronal cells that range from few hundred Pa (e.g., hydrogels) to 1–2 GPa (e.g., tissue culture polystyrene).
The PLGA membrane properties characterization was followed by its assessment in sustaining neuronal growth and differentiation. PSCD were used as reference sub-strate. This preliminary biological investigation is funda-mental to validate our membrane system as reliable hypothalamic tissue model for daidzein neuroprotective screening. To prove this, hfHypo-GnRH neurons were cultured on PLGA membranes and on PSCD and under-went to morphological observation and evaluation. SEM analysis after 3 days of culture evidenced that cells established a proper interaction at the membrane inter-face, clearly visible by the high degree of adhesion and considerable portion of membrane surface covered by the cells (Figure 1A). On both kind of substrate, PLGA mem-branes (Figures 1A–C) and PSCD (Figures 1D–F), cells reached a structural maturation, acquiring a typical bipo-lar shape, where the two deriving unbranched processes are the connection to the other cells. At first glance, the cellular reorganization over the two substrates is quite similar, thus evidencing that the PLGA membrane sys-tem is comparable to the reference one.
GnRH–secreting neurons were stained for typical neuronal lineage markers, to further assess also the
ability of the membrane system in addressing and boo-sting a full cellular differentiation. Images reconstruction was performed through a confocal microscope analysis (Figure 1B,C,E,F), which allowed to visualize the co-localization of distinctive neuronal markers, such as β-TubIII and NeuN. Representing one of the main neuro-nal cytoskeletal structural protein, β-TubIII is properly visible within the soma and along neuronal processes as indicator of an appropriate cell arrangement within the new microenvironment. NeuN was also present in the nuclei and perinuclear cytoplasm. The contemporary presence of these distinctive markers confirmed that neu-ronal cells went through the typical morphological changes of differentiation including axon sprouting and elongation, demonstrating also the good performance of the membrane system (Figure 1B). For a complete cellu-lar analysis overview, a further immunocytochemistry investigation was carried out to characterize the neuronal hypothalamic phenotype, through the identification of the positive expression of specific indicators. To assess their nature and to confirm that they are GnRH neurons, hfHypo cells were simultaneously stained for KISS1R, the receptor for kisspeptin to which hypothalamic cell source are able to respond, and for the hormone GnRH. As reported in Figure 1C, confocal images showed the coexpression of these two markers corroborating that cultured cells have a GnRH-neuron phenotype. This evidence is also sustained by the positive immuno-localization for the sex steroids receptors ER-β
F I G U R E 1 PLGA
membrane system boosts hypothalamic GnRH neuronal growth. (A,D) SEM images at different magnification and (B–C,E–F) confocal laser micrographs of neuronal cells at DIV3 in PLGA membrane system (A–C) and in PSCD control system (D–F). Cells were stained forβ-Tub ΙΙΙ (green), NeuN (red), GPR30 (magenta), KISS1R (yellow), GnRH (green), ERβ (red), and nuclei (blue). Scale bar in Confocal images is 20μm
(Figure 1B) and GPR30 (Figure 1C). Both receptors are extensively visible being typically expressed by this pecu-liar family of neurons.
As expected the PSCD, being the extensively used ref-erence system in cell culture studies, allowed the acquisi-tion of a neuronal hypothalamic phenotype thus hfHypo cells expressed all investigated neuronal markers. In con-trast, cells on PSCD were not able to express the estro-genic receptors, especially GPR30, at the same extent of those displayed on PLGA membrane system, as shown in Figure 1E, and the expression and distribution of GnRH (Figure 1F) was not wide as the one observed on PLGA membranes (Figure 1C). These evidences validated PLGA membrane system as reliable hypothalamic tissue model able also to boost the GnRH phenotypic peculiarity.
3.2
|
Modeling neuroinflammation in
the PLGA membrane system
Before assessing the neuroprotective effect of daidzein, it was necessary to set up an efficient disease model, where a realistic inflammatory neurological response evoked by a high-fat dietary intake was reproduced.
SFA palmitate was identified to play an important role as a mediator involved in the inflammation events that occur in obesity. Therefore, hypothalamic cells were exposed to different concentrations of palmitate. To iden-tify the amount of palmitate that produced a 50% reduc-tion in cell viability, cells were exposed to a concentration range of the fatty acid comprised between 0 and 600μM and the effects evaluated at the time point of 24 and 48 h (Figure 2). After 24 h, cells exposed to the concentration of 100μM of palmitate were still viable (98.6% ± 5%) with a profile quite similar to the control (100% ± 3.2%). Incubation with higher concentration of palmitate affected cell viability in a dose-dependent man-ner. A viability reduction of 52.9% ± 3.3% was observed for the concentration of 300μM with further decline when cells were treated with higher concentration (400–600 μM). The prolonged exposure for 48 h reveals palmitate toxic effect already at the lowest concentration of 100μM and a viability decrease to 55.2% ± 12%. A drastic reduction of cell viability to 6.5% ± 4%, at this time frame, is observed at the concentration of 300μM, and this trend is kept constant with the higher concentra-tion. These results highlighted cell response to palmitate time exposure, which significantly affected cell survival. The concentration of 300μM for 24 h was identified as the reference one, and used to model in vitro neu-roinflammation and to assess the neuroprotection medi-ated by daidzein co-exposure.
3.3
|
Daidzein protects GnRH cells
against palmitate-induced cell death and
reverses its inflammatory induced damage
To evaluate whether daidzein alone is toxic and influ-ences neuronal cell viability, hfHypo cells were treated with various concentrations of daidzein (0–20 μM) for 24 h. As shown in Figure 2B, cells maintained high via-bility levels demonstrating that daidzein is not toxic for neuronal cells.
The first screening of daidzein neuroprotective effects included GnRH neurons morphological observation and their viability assessments, comparing cell response to different treatments: basal condition with culture media,
F I G U R E 2 Viability of hfHypo GnRH neurons in the PLGA membrane system. (A) Dose-dependent toxic effects of palmitate on GnRH neuronal cells. Cell viability after 24 and 48 h of treatment with different concentrations of palmitate. The values expressed as average ± SD are the means of five experiments and data
statistically significant were evaluated according to ANOVA followed Bonferroni t-test (p < 0.05). * versus palmitate 200–600 μM; ǂ versus palmitate 300–600 μM; § versus palmitate 400–600 μM; # versus palmitate 600 μM; † versus palmitate 100–600 μM. (B) Cell viability after 24 h of treatment with various concentrations of daidzein
exposure to 300μM palmitate, cells co-treatment of pal-mitate (300μM) and daidzein at different concentration (0.1–1–10 μM).
SEM images reported in Figure 3A display that hfHypo cells kept in basal condition (I) adhered to the membrane and maintained their typical morphology; the induced damage, by exposing cells to 300μM palmitate (II), caused cell loss and the few cells still adhered over
the membrane surface exhibited a shrunken morphology with no projection and therefore a compromised neuro-nal cells communication. A reversed situation was visible when cells underwent to a co-treatment with the fatty acid and the daidzein (III-IV). Indeed, the acute palmi-tate treatment appeared totally neutralized by the simul-taneous exposure to daidzein as demonstrated by the wide membrane surface area covered by the cells while keeping a proper morphology of healthy cells. The same trend was observed in terms of cell viability (Figure 3B) with the graph that clearly shows an increasing cell sur-vival as increases the concentration of daidzein. When the phytoestrogen is used at the concentration of 1 and 10μM the percentage of viable cells was 89% ± 7% and 101% ±6%, respectively, which were significantly higher than that obtained in cells insulted with palmitate (58% ± 5%). At the concentration of 10μM, the viability per-centage was totally restored with value equal to the con-trol (100% ± 1%), indicating a full protection offered by the daidzein.
The obesity associated inflammation process, evoked in the GnRH cells through palmitate exposition and the cellular counteracting response, activated by the daidzein treatment, was investigated by quantifying the release of IL-6 and IL-10, reported in the graph in Figure 4A,B, respectively. Neuropathological changes are often associ-ated with the increased IL-6 expression in brain, since this cytokine plays a pivotal role in the inflammatory response. Figure 4A depicts IL-6 release and as expected the palmitate treated cells delivered the highest amount of the pro-inflammatory cytokine (274 ± 26 pg/106cell) as a sign of the ongoing neuroinflammation process. The IL-6 production is gradually reduced when the palmitate inflammatory stimulus is accompanied by the co-treatment with daidzein. The graph highlights a progres-sive IL-6 reduction as the concentration of daidzein increases; the inhibition of IL-6 production is quite signif-icant (p < 0.05) at the daidzein treatment of 10 μM as demonstrated by the reduction of IL-6 secretion at a value of 51.9 ± 11.3 pg/106cell, which is similar to that of the untreated cells (22.2 ± 4.3 pg/106cell). To further asses the daidzein anti-inflammatory properties, the IL-10 secretion was also analyzed. The anti-inflammatory cytokine IL-10 is responsible for balancing the immune response in the brain and therefore produced as a defense reaction to an inflammatory mediator. The ability of daidzein to reverse neuroinflammation was supported by the evidence in graph 4B in which the only palmitate treatment induces a peak of IL-10 (148.5 ± 23 pg/106cell), but its production is consistently attenuated by daidzein co-treatment in a concentration dependent manner.
The high IL-6/IL-10 ratio measured in the palmitate-treated cells (1.87 ± 0.34) decreased as the concentration
F I G U R E 3 Daidzein protected GnRH cells against palmitate-induced cell death. (A) Representative images depicting the morphology of neuronal cells after 24 h of treatment with Palmitate (300μM) and Daidzein at different concentrations; Control cells without any treatment (I); Palmitate (300μM) treatment (II); co-treatment with: palmitate and daidzein 1μM (III), palmitate and daidzein 10μM (IV). (B) Viability of GnRH cells after treatment with of Palmitate (300μM) and Daidzein at different
concentrations (0.1–1–10 μM) for 24 h. The values are presented as percentages of control (no palmitate treatment). Data are mean ± SEM from three independent experiments and data statistically significant were evaluated according to ANOVA followed Bonferroni t-test (p < 0.05): * versus other treatments; ǂ versus palmitate 300μM and Palmitate+ daidzein 0.1 μM
of daidzein increases, reaching value equal to the control (0.69 ± 0.47) when cells were treated with daidzein 10 μM (0.63 ± 0.17) (Table 2). These findings indicate that the overall inflammation process was silenced in response to the daidzein treatment.
3.4
|
Daidzein neuroprotection is
mediated by its antioxidant activity and
anti-apoptotic effects
We explored if the neuroprotection, offered by daidzein to palmitate insulted cells, is evoked by its antioxidant effect. Therefore, to disclose this mechanism we proceeded to seek the ability of daidzein in preventing intracellular ROS accumulation. To this purpose, the H2DCFDA probe was used to highlight the redox balance
maintenance through daidzein activity. As shown in Figure 5A palmitate treatment led to a massive produc-tion of ROS as qualitatively evidenced by the bright green fluorescence compared with the control. Quantitative analysis of the green fluorescence emission, whose inten-sity is directly linked to the ROS production and accumu-lation, showed that daidzein co-treatment significantly reduced ROS production in a dose-dependent manner (Figure 5B). The stronger intracellular antioxidant activ-ity was obtained at the daidzein concentration of 1 and 10μM, quantifying a ROS intracellular in the same physi-ological range of normal cell metabolism.
To further characterize daidzein neuroprotective properties, its action in preventing the apoptotic cas-cade was also analyzed. To fulfill this task as first step was investigated the phytoestrogen potential to avoid mitochondrial dysfunction. Since mitochondria are inherently involved in the apoptotic process of cells, the JC-1 dye was used to monitor mitochondrial per-meability; indeed, damaged cells undergo to a progres-sive pore opening with the consequent loss of the mitochondrial permeability potential (MMP), which is an intracellular event that occurs during apoptosis. The JC-1 staining was evaluated in GnRH neuronal cells kept in different culture conditions in order to assess if daidzein could protect cells from the palmitate-induced mitochondrial damage. As shown in Figure 6A when cells are treated only with palmitate (II) and co-treated with daidzein 0.1μM (III) the green fluorescence signal increased, compared with the control and to the other samples co-treated with other concentrations of daid-zein. The increase of the green fluorescence emission
F I G U R E 4 Anti-inflammatory role of Daidzein. Daidzein suppressed palmitate-induced cytokine secretion from GnRH neurons: IL-6 (A) and IL-10 (B) production. The values expressed as average ± SD are the means of three experiments and data statistically significant were evaluated according to ANOVA followed Bonferroni t-test (p < 0.05). * versus other treatments; ǂ versus control, palmitate + daidzein 1 μM, palmitate + daidzein 10μM; † versus control and palmitate + daidzein 10 μM; § versus control
T A B L E 2 IL-6, IL-10, and IL-6/IL-10
ratio Treatment IL-6 [pg/10
6
cell] IL-10 [pg/106cell] IL-6/IL-10 ratio
Control 22.2 ± 4.3 29.2 ± 18 0.69 ± 0.47
PA 274 ± 26 148.5 ± 23 1.87 ± 0.34
PA+ DZ 0.1μM 212.6 ± 14.8 121.9 ± 22 1.79 ± 0.39
PA+ DZ 1μM 97.6 ± 15.4 95.4 ± 21 1.02 ± 0.28
PA+ DZ 10μM 51.9 ± 11.3 68.9 ± 23 0.63 ± 0.17
Note: Data are presented as mean ± SD. Abbreviations: DZ daidzein; PA, palmitate.
indicates mitochondrial damage on the basis of the JC-1 dye mechanism: it enters the cells from the cyto-sol to the mitochondria and forms aggregates which emit a red fluorescence while in apoptotic cells, due to mitochondrial membrane leakage, JC-1 aggregates go into the cytosol in monomeric form which emits green fluorescence. The red fluorescence signal increased by increasing daidzein concentration as visible in the rep-resentative confocal images (Figure 6IV–V) proving that the investigated molecule prevented the mitochon-dria damage. Quantification of the fluorescence inten-sity ratio between red/green channels (Figure 6B), resulting from the proportion healthy/damaged mito-chondria, reveals how palmitate totally disrupts the mitochondrial membrane integrity and decreases the MMP. The ratio between red/green fluorescence signal increases as an effect of the diadzein exposure, with a rescue activity more evident at the daizein higher con-centration. Daidzein prevented MMP collapse exerting a decisive neuroprotection, especially at the concentra-tion of 10 μM the palmitate side effects are totally
counterbalanced. Next, we investigated a more specific aspect of apoptotic process through the immuno-staining for apoptotic markers by localizing and quanti-fying the presence of positive cell for Caspase 3 and p-JNK (Figure 6C). After palmitate treatment 38% ± 13% and 18% ± 3% of the cells are positive for Caspase 3 and p-JNK, respectively, acting as pro-apoptotic cell death inducers. In line with the previous evidence, the co-treatment with daidzein reversed the trend and although cells were still under the effect of 300μM pal-mitate, the synergistic treatment with daidzein inhibited the palmitate neuro-toxic effects and the related cells death. Indeed, the number of cells positive for the two markers significantly decreased, obtaining at the daidzein highest concentration (10μM) values of 3% ± 1% and 2% ± 0.7%, for caspase 3 and p-JNK, respectively, which correspond to the same positive cell count observed in the control. Disclosing the anti-apoptotic effects of daidzein was highlighted another action through which the phytoestrogen exerts neuroprotection.
F I G U R E 5 Anti-oxidant effect of Daidzein. Daidzein inhibited palmitate-induced ROS generation in neuronal cells. (A) Confocal laser micrographs displaying DCF fluorescence signal in hypothalamic GnRH neurons in basal condition (Control) (I), after palmitate treatment (300μM) (II) and after co-treatment with palmitate and daidzein 0.1 (III), 1 (IV) and 10 μM (V). Scale bar: 20 μm. (B) Quantitative analysis of DCF fluorescence intensity produced in neuronal cells incubated with various concentrations of daidzein and palmitate. The values expressed as average ± SD are the means of five experiments and data statistically significant were evaluated according to ANOVA followed Bonferroni t-test (p < 0.05): * versus other treatments; ǂ versus control, palmitate + daidzein 1 μM, palmitate + daidzein 10 μM
3.5
|
Daidzein prevents the
palmitate-induced hfHypo cell dysfunction
As previously described, in standard culture conditions hfHypo cells widely express markers of the neuronal line-age, such asβ-Tub III and NeuN, as a proof of a proper degree of differentiation. The GnRH phenotype is also confirmed by the expression and cellular distribution of the two specific markers GnRH and KISS1R and by the positive staining for the estrogenic receptors ERβ and GPR30 (Figure 7A). The capacity of daidzein to provide neuroprotection was also investigated by assessing its mediated effect to preserve the cellular structural integ-rity under the palmitate inflammatory stimulus. Palmi-tate toxicity induced a consistent reduction in cell number and the complete disruption of the normal struc-ture organization with a drastic reduction in the expres-sion of the neuronal marker β-Tub III and NeuN (Figure 7A). In damaged hfHypo cells, the GnRH pheno-type markers, GnRH and KISS1R are weakly expressed. The estrogenic receptors ERβ and GPR30 are also visible only in a small fraction of cells; thus, palmitate evoked its neurotoxicity through a massive neuronal dedifferentia-tion and impacting the funcdedifferentia-tional integrity of the GnRH secreting cells. A complete protection was offered to the cells by the co-treatment with daidzein 10 μM; the pres-ence of the phytoestrogen totally compensated the palmi-tate damage. Confocal pictures reveal healthy cells expressing NeuN and with a uniform distribution of the β-Tub III along the unipolar neurite elongation. GnRH peptide, KISS1R and the estrogenic receptors are broadly
F I G U R E 6 Anti-apoptotic effect of daidzein. (A) Daidzein protected GnRH neurons against palmitate-induced mitochondrial membrane depolarization. Representative confocal images of neurons displaying JC1 monomers (green) and aggregates (red) in basal condition (Control) (I), after palmitate treatment (II) and after co-treatment with palmitate (300μM) and daidzein 0.1 (III), 1 (IV), and 10μM (V). (B) Quantitative analysis of the fluorescence intensity ratio between JC-1 red and green. The values expressed as average ± SD are the means of five experiments and data
statistically significant were evaluated according to ANOVA followed Bonferroni t-test (p < 0.05): * versus other treatments; ǂ versus palmitate and palmitate + daidzein 0.1 μM treatments. (C) Daidzein inhibited palmitate-induced apoptotic markers activation. Quantitative analysis of apoptotic markers Caspase-3 (green bar) and p-JNK (red bar). The values expressed as average ± SD are the means of five experiments and data statistically significant were evaluated according to ANOVA followed Bonferroni t-test (p < .05): * versus other treatments; ǂ versus control, palmitate + daidzein 1 and 10μM
expressed evidencing that daidzein prevented palmitate toxicity protecting cells from losing the peculiar pheno-type markers. The profile expression patterns for each of the above-mentioned markers was also quantified and expressed as the percentage of positive cells over the total number of cells within the investigated image field (Figure 7B). After palmitate treatment, the percentage of cells positive to the neuronal markers β-Tub III and
NeuN, is of 80% ± 6% and 53% ± 9%, respectively, which are values significantly lower compared with the control. The number of cells expressing the neuronal lineage pro-teins increases to 97% ± 3% for β-Tub III and to 88% ± 9% for NeuN when the fatty acid is administered together with daidzein. This trend is kept constant and even emphasized for the other markers, where the per-centage fluctuations of positive cells are more evident. In
F I G U R E 7 Daidzein protected GnRH neurons against palmitate-induced cell damage. (A) Confocal laser micrographs of neuronal cells in PLGA membrane system after different culture conditions: cells without any treatment (Control), Palmitate treatment (Palmitate), and co-treatment with palmitate and daidzein 10μM (PA+ DZ). Distribution pattern profile of neuronal markers β-Tub ΙΙΙ (green), and NeuN (red), GnRH phenotype markers KISS1R (yellow) and GnRH (green), estrogenic receptors ERβ (red) and GPR30 (magenta). Nuclei are
counterstained with DAPI (blue). Scale bar: 20μm. (B) Quantitative analysis of positive cells for each marker. Results are expressed as percentage of positive cells over total and are reported as average ± SD of the means of five experiments. Data statistically significant were evaluated according to ANOVA followed Bonferroni t-test (p < 0.05): * versus palmitate treatment
Figure 7B is clearly visible a drastic downregulation of cells expressing GnRH (20% ± 6%), KISS1R (22% ± 7%), ERβ (18% ± 7%), and GPR30 (25% ± 4%) in response to palmitate insult. In presence of daidzein the percentage of hfHypo positive cells for these specific proteins signifi-cantly increases to values of 92% ± 2%, 72% ± 6%, 81% ± 7%, and 84% ± 6%, respectively, in the same range of the control. Daidzein suppresses the induced palmitate cell disfunctions, and the inhibitory palmitate effect over the GnRH phenotype markers and estrogenic receptors expression is not detectable anymore. Thanks to the daidzein-mediated neuroprotection the markers expres-sions remain within a physiological range as a proof of an appropriate cell functionality.
4
|
D I S C U S S I O N
Overnutrition-induced hypothalamic inflammation due to high-fatty acid diet has become a serious public health problem because it is associated with severe con-sequences, including obesity, diabetes, cardiovascular and neurodegenerative diseases. Thus, many efforts are directed to identify the compounds or molecules that can inhibit or reduce the activation of inflammatory signaling in hypothalamic neurons in order to prevent or treat obesity and related diseases. The challenge is to stop the progressive neurodegeneration by inducing neuroprotection.
The present study provided novel insights on the anti-neuroinflammatory effects of phytoestrogen daidzein, potentially relevant to the understanding of the neuro-protective role of this compound in modulating the inflammatory processes implicated in the evolution of several chronic diseases.
Innovative investigational insights on new and effec-tive therapeutics to treat neurological disorders, and cor-related pathological events, require the setting up of a reliable in vitro platform in which the in vivo structure complexity is recapitulated and maintained.
The in vitro reconstruction of GnRH hypothalamic tissue-specific microenvironment, created in this work by using PLGA membrane system, served as model system to test daidzein's ability to stop and reverse the imbal-anced neuroinflammatory responses caused by high fat diet.
PLGA membranes, thanks to their morphological, mechanical and physico-chemical properties, enable adhesion, growth and structural and functional neuronal differentiation of the hfHypo-GnRH neurons, by provid-ing a microenvironment well-controlled at molecular level.
In line with a previous study that reports fetal hypo-thalamus as a rich source of GnRH neurons,23 hfHypo cells retain their GnRH phenotype in vitro within the PLGA membrane system. Indeed, they displayed a unipo-lar or bipounipo-lar morphology with long processes to connect each other. Immunohistochemical data corroborated the GnRH neuronal phenotype of hfHypo cells, which, besides markers of the neuronal lineage, such as β-Tub III and NeuN, also abundantly express GnRH peptide and the functional kisspeptin receptor, KISS1R, as well as the estrogenic receptor ERβ, and the transmembrane estrogen receptor, GPR30, which has been recently impli-cated in rapid action of estrogen in GnRH neurons.27
The neuronal biohybrid membrane system developed in this work, consisting of the hfHypo-GnRH neurons in culture on the PLGA membrane, reproduces the morpho-functional characteristics of neuronal tissue, and repre-sents, therefore, a valid model-system of hypothalamic tissue. The possibility to set up a realistic biomimetic platform allows to use this system for a breakthrough investigation of the pathogenic mechanisms at the basis of the fatty acids/diet neuroinflammation, and also for the screening of bio-molecules to evaluate their pharma-cological potential. With this aim, the PLGA membrane hypothalamic tissue model was used to induce neu-roinflammation associated with obesity, in order to inves-tigate the potential anti-neuroinflammatory effect of phytoestrogen daidzein.
To mimic the inflammatory process associated with obesity, cultured cells in the PLGA membrane system were treated with palmitate, a SFA widely present in the diet, whose excessive intake may induce a hypothalamic inflammatory state, as demonstrated by several studies in the literature and confirmed by the present work. The toxic effect of palmitate involves an increase of pro-inflammatory cytokines, elevated oxidative stress, and can result in the death of neuronal cells by apoptosis via a JNK-dependent pathway.3,25
We have demonstrated, for the first time in primary cultures of hypothalamic GnRH neurons of human ori-gin, that the proinflammatory events induced by palmi-tate treatments have involved the increase of inflammatory cytokines, reduction of cell viability, changes in morphology, oxidative stress, and cell death by apoptosis. Moreover, palmitate action also implicates the reduction of specific cellular functions by inhibition of the expression of neuronal markers as well as the GnRH peptide, kisspeptin receptor (KISS1R) and estro-genic receptors (ERβ and GPR30). Then, we evaluated whether daidzein may modulate the deleterious effects triggered by palmitate administration and related to obesity-induced neuroinflammation in primary cultures
of human GnRH neurons by using the developed neural membrane system that mimics the in vivo conditions. In particular, the study characterized the effects of daidzein on palmitate-treated GnRH neurons, with the aim to evaluate the potential capabilities of this phitoestrogen to maintain normal cellular phenotype and protect from oxidative damage, apoptosis, and the deleterious effects of palmitate-induced inflammation on GnRH neuronal functions.
The morphology of neuronal cells reflects their struc-tural and functional integrity, and therefore represents an index of neuronal growth, viability, and differentia-tion. We observed how palmitate treatment induced a significant decrease of the number of cells and morpho-logical changes of GnRH neurons that assumed a round shape of the cellular body with high neurite loss, which is one of the hallmarks of neuronal injury and neu-rodegeneration. It is evident, therefore, that the palmitate caused a cellular damage capable of modifying the mor-phological characteristics typical of such kind of neurons, which are indispensable to accomplish the normal physi-ological cellular functions.
Daidzein restored the morphological characteristics observed in the control cells, consisting of untreated cells, and furthermore, induced an increase in cell viability in a concentration-dependent manner highlighting the neuro-protective action of daidzein against the toxicity induced by palmitate. These findings are consistent with previous studies that showed that the administration of daidzein caused increased survival of cortical neurons exposed to oxygen and glucose deprivation,12significantly promoted the vitality and proliferation of hippocampal neurons and stimulate the neuritogenesis process.28
To elucidate the anti-inflammatory mechanism induced by daidzein, further investigations were carried out focusing on the involvement of cytokines production, such as IL-6 and IL-10 that have emerged as key regula-tors of neuroinflammation. IL-6 is classified as a pro-inflammatory cytokine that may have an anti-inflammatory effect negatively regulating the inflamma-tion of acute phase response. In models of brain injury, infection, LPS injection, or diet-induced obesity, IL-6 is upregulated and modulates inflammation, apoptosis, and oxidative stress. Alternatively, IL-6 is also involved in dampening of the immune responses to promote recovery and healing.24
IL-10 is an essential anti-inflammatory cytokine that is released from several cell types, after immune chal-lenge and brain injury, and is responsible for tissue remo-deling and restoration of homeostatic conditions. Barabas et al., provided the first evidence that IL-10 directly regu-lates the functions of GnRH neurons.29
Daidzein suppressed palmitate-mediated production of the investigated inflammatory cytokines demonstrat-ing that daidzein had also direct anti-inflammatory prop-erties in addition to its function to prevent cell death. These results are in agreement with a previous report, in which daidzein treatment significantly attenuated the inflammatory IL-6 expression in an in vitro model of adi-pose inflammation,13 suppressed IL-10 secretion from LPS treated human dendritic cells,30 and attenuated cisplatin-induced inflammatory cytokines interleukin 10 (IL-10) in mice with renal damage.9To the best of our knowledge, this is the first study showing that daidzein inhibits pro-inflammatory cytokines in human GnRH neurons. Additionally, Ropelle et al., report that recombi-nant IL-6 requires IL-10 expression to suppress hyperpha-gia in the hypothalamus of obese animals.31IL-6 induces an anti-inflammatory environment by inducing the pro-duction of IL-10. Therefore, we can hypothesize that palmitate-induced increase of IL-10 levels may be related to the IL-6 anti-inflammatory response, which involves the increase of IL-10 expression in the hypothalamus, accordingly with Ropelle.31 In addition, to evaluate the balance between pro- and anti-inflammatory cytokines which controls various physiological processes and the imbalance of which is related to several human diseases, the decrease of IL-6/IL-10 ratio obtained administering daidzein in palmitate-treated cells demonstrates its role in inhibiting the inflammation process.
It is noteworthy that cytokines had an effect on gene expression in GnRH neurons or neuron function, since GnRH neurons express cytokine receptors.32 Indeed, we previously demonstrated deleterious effects of TNFα on KISS1R signaling and GnRH secretion in the same cellu-lar model.23 In the present work, we found an effect of palmitate-induced inflammation on cytokines production and phenotype markers, as well as estrogenic receptors. In particular, the palmitate inflammation induced an increase of IL-6 and IL-10 and a decrease of KISS1R, GnRH, ERβ, and GPR30 expression highlighting a rela-tionship between cytokines and markers of phenotype and estrogenic receptors, in accordance with previous studies.23,32This effect was completely reversed by daid-zein treatment that plays a neuroprotective role in impaired hfHypo cells.
Particular attention was paid to the ability of phytoes-trogen daidzein to counteract the negative and toxic effects caused by palmitate on phenotype and function of hypothalamic GnRH neurons. Palmitate treatment cau-sed a very low expression and localization of typical neu-ronal lineage markers such asβ-Tub III and NeuN that leads to the loss of their well-defined neuronal morphol-ogy. The SFA exposure determined a specific
inflammatory intracellular signaling characterized by reduced GnRH and KISS1R expression.
This direct inhibitory effect on KISS1R expression observed in our study is consistent with recent reports that demonstrated the association of impairment caused by inflammatory injury with a downregulation of KISS1R in GnRH neuron.23,33Our finding supports the idea that the deleterious effects of inflammation on GnRH neurons function, activated by palmitate, are exerted through the decrease of GnRH and KISS1R expression. In line with our expectations, the administration of palmitate was able to decrease not only the number of neuronal cells, but it also negatively affected the phenotype and function of hypothalamic GnRH-secreting neurons.
Daidzein restored the neuronal morphology and the correct GnRH phenotype. The intense distribution of β-Tub III and NeuN markers indicated the neuronal phe-notype of primary culture. Daidzein significantly inhibited the adverse effects of palmitate on GnRH expression, accordingly with in vivo data that showed how the administration of daidzein determined a marked up-regulation of the levels of mRNA of GnRH in the hypothalamus.34 Accordingly, daidzein-treated hfHypo cells exhibited the predominant hypothalamic GnRH/
KISS1R phenotype, as demonstrated by
immuncytochemical analysis. The ability of daidzein in restoring both GnRH and KISS1R expression are extremely relevant by a functional point of view, since kisspeptin is the main physiological regulator of the hypothalamic GnRH system, which governs the repro-ductive function.
The protective actions of phytoestrogens are mediated by the activation of intracellular nuclear receptor family, ERα and ERβ, which are located in the cytoplasm or on the nuclear membrane. Recent studies have suggested that the transmembrane G protein-coupled receptor 30 (GPR30) may be involved in estrogen anti-inflammatory action. Using human fetal primary cholin-ergic neurons, it has been recently demonstrated that GPR30 is specifically implicated in mediating neuro-protective effects of estrogens under inflammatory stimu-lus.27 Other investigations report that daidzein in the presence of glutamate, used to induce toxicity in primary hippocampal neurons, decreased the level of ERβ without significant changing of the GPR30 protein level.35 Our results showed, for the first time that daidzein is able to increase the expression levels of both ERβ and GPR30 in the presence of palmitate-induced inflammatory stimulus in hypothalamic GnRH neurons, thus suggesting that these receptors could be involved in the mechanism of action of daidzein in determining neuroprotection against neuroinflammation of obese individuals. In sup-port of this hypothesis, scientific data showed that
daidzein, thanks to its structural homology with 17 β-estradiol, is able to bind to the classical estrogenic recep-tor ERβ in hippocampal neurons, stimulating the growth of their axons36; while more recent data suggest that its main metabolite equol is able to inhibit neu-roinflammation induced by LPS in astrocytes by pathway mediated by the GPR30.37
SFA induce mitochondrial damage and oxidative stress that are accompanied by excessive production and/or insufficient removal of ROS. In general, isoflavones are known to be antioxidant, and, indeed, daidzein suppressed ROS production in microglia during LPS-induced neuroinflammation,8 and also in mouse models of cerebral ischemia reduced oxidative stress cau-sed by increacau-sed activity of the enzyme superoxide dis-mutase.38The neuroprotection effect of daidzein could be due to its chemical structure that allows to play a role of scavenger of free radicals by hydrogen atom donation, protecting in this way cells from oxidative damage. Herein, we have demonstrated that the treatment with daidzein avoids palmitate-induced ROS production that is responsible for cell impairment. These results clearly indicate and thus confirm the antioxidant properties of such phytoestrogen.
It is known that the neuroinflammation associated with obesity is characterized also by the apoptotic process activation. Excessive ROS, induced by palmitate, can lead to mitochondrial dysfunctions with loss of the mitochon-drial membrane potential (MMP), which is widely con-sidered as an early intracellular event of the apoptotic cascade. Palmitate caused a drastic depolarization of MMP and the treatment with daidzein had a concentration-dependent protective effect on mitochon-dria. Therefore, our results demonstrated that daidzein can prevent the palmitate-induced collapse of MMP, suggesting a strong protective effect against palmitate-induced mitochondrial dysfunction. Thus, daidzein neu-roprotective action could be related to the recovery of mitochondrial bioenergetics and functions, and the capacity to inhibit the apoptotic process triggered by pal-mitate, suggesting therefore its anti-apoptotic effect.
Several molecular studies provided evidences that inflammatory responses are mediated by JNK activa-tion.39Recently, Tse et al.25demonstrated that activation of JNK is required for palmitate-induced inflammation, specifically, other authors report that SFA palmitate, may induce neuroinflammation through pathways involving the activation of JNK and caspase 3.2Moreover, daidzein inhibited secretion of pro-inflammatory cytokines in adipocyte-macrophage co-cultures via multiple pathways in which JNK activation is involved.13
Within this scenario, we further explored the daid-zein anti-apoptotic effect by investigating the ability of
this phytoestrogen in preventing palmitate-induced acti-vation of JNK and Caspase-3, since they are implicated in the sequence of events leading to apoptosis. Indeed, JNK represents the upstream sensor protein, which is activated following a condition of oxidative stress due to palmitate treatment, and caspase 3 represents the downstream apoptotic effector molecule. To have a gen-eral overview at upstream and downstream of JNK acti-vation, we have evaluated the capacity of daidzein to inhibit ROS production, since it acts as potent inducer of JNK, and one of the antiapoptotic mechanisms is to remove the ROS to prevent JNK activation.40 JNK is also involved in other apoptotic signal activation as the downstream mitochondrial/caspase-9 pathway whose activity converges on the activation of caspase-3, which is then a downstream effector of JNK pathway.41 Importantly, the treatment with daidzein blocked JNK phosphorylation and caspase-3 cleavage and, thus, prevented the deleterious effects of palmitate by regulat-ing signalregulat-ing pathways within neurons and inhibitregulat-ing the onset of the apoptotic process. The study also
analyzed the loss of mitochondrial membrane potential that is considered as an early indicator of apoptosis. Therefore, we sought for representative events that occur in the apoptotic cascade, with the final goal to provide evidence about daidzein's anti-apoptotic effect, to gain insight into the mechanism underlying the daidzein-induced neuroprotection.
Overall, our data demonstrated that cell death, mor-phological changes, interleukins and ROS production, as well as apoptotic state, and impaired functionality evoked by palmitate-activated neurons, were significantly inhibited in a dose-dependent manner by daidzein co-treatment. The putative mechanism of the neuro-protetctive effect of daidzein against palmitate-induced hypothalamic inflammation is summarized in Figure 8.
On the basis of the collected data we can suppose that daidzein, thanks to its chemical structure, which includes a phenolic ring, can directly bind to the estrogen recep-tors and regulate the hypothalamic cellular functions, coherently with other studies.42 Overall, the present study highlights the therapeutic potential of daidzein in
F I G U R E 8 Anti-neuroinflammatory effect of daidzein in GnRH neurons. Daidzein was able to counteract the detrimental effects of palmitate by eliciting an increased viability and diminishing ROS and interleukins production. Daidzein displayed anti-apoptotic action by downregulating mitochondrial dysfunction and p-JNK and caspase-3 activation. Importantly, daidzein upregulated markers of neuronal linage and GnRH phenotype and also mediated the activation of ERβ and GPR30 receptors in palmitate-treated cells
the prevention of neuroinflammation caused by high-fat diets.
5
|
C O N C L U S I O N S
This study provides novel insight into the anti-neuroinflammatory effects of daidzein on hfHypo cells. PLGA membrane system enhanced the creation of func-tional hypothalamic neuronal tissue analogue that was successfully employed as a valuable tool for testing the neuroprotective effect of daidzein against high fat diet-induced inflammation.
The molecular mechanism of daidzein neuroprotection was mediated through downregulation of palmitate-induced cell impairment. Specifically, the phytoestrogen not only inhibited palmitate-induced neuronal death and pro-inflammatory cytokines secre-tion, but also protected from neuroinflammatory injury through downregulation of ROS production, mitochon-drial dysfunction and p-JNK and caspase-3 activation, thus demonstrating both antioxidant and anti-apoptotic action. Furthermore, daidzein was able to suppress the inhibitory effect of palmitate on the expression of specific markers of neuronal linage and GnRH phenotype, suggesting its role in restoring the correct neuronal morphology and functionality of hfHypo GnRH cells. In addition, its protective action was also mediated by the activation of ERβ and GPR30 receptors.
Our data highlight the therapeutic potential of daid-zein in counteracting neuroinflammation caused by high-fat diets, and therefore, in the prevention of the related diseases. Since GnRH neurons play a key role in controlling the hormonal network crucial for sus-taining the reproductive function, our findings could advance the understanding not only of the mechanisms underlying daidzein neuroprotective effect, but also of the close relationship between obesity-related neu-roinflammation/neurodegeneration and reproductive function.
AKNOWLEDGMENTS
The study was supported by a grant from Ministero dell'Università e della Ricerca, Italy (MIUR-PRIN2017, grant no. 2017TK7Z8L_006) to G. G. and A. M. The authors acknowledge Mariano Davoli, Department of Biology, Ecology and Earth Science, DiBEST, University of Calabria, Italy, for providing technical help with SEM analysis.
C O N F L I C T O F I N T E R E S T
All authors declare that there are no conflicts of interest. A U T H O R C O N T R I B U T I O N S
Sabrina Morelli: conceptualized, designed and carried out experiments, analyzed results and wrote the manuscript. Antonella Piscioneri: prepared mem-brane system, performed cell culture experiments, immunostaining, contributed to analyzing and writ-ing results. Giulia Guarnieri: contributed to cell cul-tures; Annamaria Morelli: contributed to the interpretation of the results and revised the manu-script. Enrico Drioli: revised the manuscript. Loredana De Bartolo: designed experiments, analyzed results and revised manuscript. All authors read and approved the final version of the manuscript.
D A T A A V A I L A B I L I T Y S T A T E M E N T
The data that support the findings of this study are avail-able from the corresponding author upon reasonavail-able request.
O R C I D
Sabrina Morelli https://orcid.org/0000-0001-7551-8427
Antonella Piscioneri https://orcid.org/0000-0002-1052-1550
Giulia Guarnieri https://orcid.org/0000-0002-3056-7492
Annamaria Morelli https://orcid.org/0000-0001-8027-9870
Enrico Drioli https://orcid.org/0000-0003-1092-5368
Loredana De Bartolo https://orcid.org/0000-0003-3774-9114
R E F E R E N C E S
1. Cai D. Neuroinflammation and neurodegeneration in
overnutrition-induced diseases. Trends Endocrinol Metab. 2013;24(1):40–47.
2. Mayer CM, Belsham DD. Palmitate attenuates insulin signaling and induces endoplasmic reticulum stress and apoptosis in hypothalamic neurons: rescue of resistance and apoptosis through adenosine 50 monophosphate-activated protein kinase activation. Endocrinology. 2010;151:576–585.
3. McFadden JW, Aja S, Li Q, Bandaru VVR, Kim E-K, Haughey NJ, et al. Increasing fatty acid oxidation remodels the hypothalamic neurometabolome to mitigate stress and inflam-mation. PLoS One. 2014;9(12):e115642.
4. Truong VL, Jun M, Jeong WS. Role of resveratrol in regulation of cellular defense systems against oxidative stress. Biofactors. 2018;44(1):36–49.
5. Das D, Sarkar S, Bordoloi J, Wann SB, Kalita J, Manna P. Daid-zein, its effects on impaired glucose and lipid metabolism and vascular inflammation associated with type 2 diabetes. Biofactors. 2018;44(5):407–417.
